Journal of Petroleum & Environmental Biotechnology

Deriving permeability models from static 3D seismic attributes through artificial intelligence tools for conventional gas reservoirs

4^th International Conference on Petroleum Engineering

August 15-17, 2016 London, UK

AbdulRahman M AlNutaifi

Saudi Arabian Oil Company, Saudi Arabia

Posters & Accepted Abstracts: J Pet Environ Biotechnol

Abstract:

Developing gas reservoirs and assets is playing major role in supplying consumers demand provided that those developments are optimized with the minimal capital cost investment. A major challenge in those reservoirs is the reservoir quality and the ability to target potential productive sweet spots. It is typical in any delineation or development program to account for potential dry holes which in some cases escalate the drilling requirements substantially to meet the delineation or development objectives. In the current reservoir characterization practices, well placement is mainly based on the derived porosity distributions from Impedance or other seismic attributes. Only actual drilling results will reveal the reservoir performance and its flow capacities. Nevertheless, the advancements in mathematical models with the introduction of Artificial Intelligence tools can help in correlating the seismic not only to porosity but also to permeability. The AI models are able to correlate the input data with the help of what is called training sets. After the model is created it is calibrated with the testing set. In this project, we will construct a reservoir permeability model from 3D seismic attributes using Artificial Intelligence tools. First, we will identify the relevant 3D seismic attributes and windows that will serve as the input data to the AI models. Secondly, several AI models which include Artificial Neural Networks, Fuzzy Systems and several types Vector Machines will be evaluated and optimized with several training and testing sets from two fields in Saudi Arabia with enough penetrations. The models will then be discussed and validated with some drilling results out of those two fields to check for validity and further recommendations for future work will be discussed.

Biography :

Email: anutaifi@gmail.com